Perioperative atrial arrhythmias in noncardiothoracic patients: a review of risk factors and treatment strategies in the veteran population

Jennifer Marye Burris,Anuradha Subramanian,Shubhada Sansgiry,Carlos H Palacio,Faisal G Bakaeen,Samir S Awad
DOI: https://doi.org/10.1016/j.amjsurg.2010.07.019
Abstract:Background: Perioperative atrial arrhythmias (PAAs) in noncardiothoracic patients have poorly defined risk factors and management. Methods: The surgical intensive care unit database was queried for patients who developed PAAs from 2008 to 2009. Demographics, comorbidities, preoperative data (electrocardiography, chest x-rays, laboratory results), medications, intraoperative variables, management, and outcomes of atrial arrhythmias were collected. Controls were randomly chosen in a 3:1 ratio. Comparisons were performed using χ² tests, Student's t tests, or nonparametric comparisons as appropriate. Multivariate logistic regression was performed. Results: Five hundred sixty-one patients were admitted to the surgical intensive care unit. Three hundred fifty-four (63%) had noncardiothoracic surgery, and 30 (8.5%) developed PAAs. The mean age of patients with PAAs was 66 ± 7.3 years, compared with 64 ± 11 years for controls (P = NS), with most patients undergoing general (60%) and vascular (33%) surgery. PAA patients were more likely to have coronary artery disease (P = .029), cardiomegaly (P = .011), and premature atrial contractions (P = .016) and to take aspirin (P = .010). On multivariate logistic regression, predictors of atrial arrhythmias were premature atrial contractions, preoperative hypokalemia, intraoperative adverse events, and cardiomegaly. Most PAA patients received amiodarone (63%). Ten percent required electrical cardioversion, and 26% received anticoagulation. PAA patients had significantly longer intensive care unit lengths of stay (P = .032). Conclusion: Coronary artery disease, cardiomegaly, hypokalemia, and premature atrial contractions were significantly associated with PAAs in noncardiothoracic patients. Prospective studies are needed to define treatment guidelines.
What problem does this paper attempt to address?